Furniture structure quality detection method and device based on image recognition

Abstract

The invention relates to the field of furniture quality detection, and discloses a furniture structure quality detection method and device based on image recognition, and the method comprises the steps: obtaining a plurality of sample furniture images of a plurality of furniture products, segmenting the sample furniture images into a plurality of sub-sample region images, and dividing the plurality of sub-sample region images into a plurality of sub-sample groups; marking the sub-sample region images in the plurality of sub-sample groups, and training a sub-detection system according to the marked sub-sample region images in the sub-sample groups; according to the furniture structure, a plurality of sub-detection systems corresponding to the furniture structure are determined, the plurality of sub-detection systems corresponding to the furniture structure are deployed, and the plurality of sub-detection systems are used for detecting the product quality of the hole site, the frame corner and the frame joint of the furniture structure. According to the invention, rapid detection of furniture quality can be carried out for rapidly developed and changed furniture product structures.

Description

Technical Field

[0001] The present application relates to the technical field of furniture quality detection, and more specifically, to a method and device for detecting furniture structure quality based on image recognition. Background Art

[0002] In the production of small furniture, the quality of furniture is generally tested through manual quality inspection to ensure that it meets specific quality standards. When checking the appearance of furniture through manual inspection, it is necessary to use the naked eye to check whether the shape meets the regulations, whether there are scratches, cracks or other defects on the surface, and even use tools such as tape measures and calipers to measure the size of the furniture to ensure that the furniture meets the design specifications. Manual inspection and evaluation of the quality of small furniture is time-consuming and laborious, and it is easy to miss the size and structure of small furniture.

[0003] With the development of technology, image recognition technology has been widely used in the field of furniture quality inspection, especially for small furniture production lines. Image recognition quality inspection improves efficiency and consistency by using cameras and computer vision algorithms to automatically detect furniture quality problems. However, the existing image recognition technology cannot adapt to the rapid changes in the structural design of small furniture when inspecting furniture quality, and cannot quickly detect the rapidly developing and changing product structure of furniture. For example, patent CN118823029B (application number: CN202411314735.1) provides a furniture surface quality detection method based on machine vision, which includes: obtaining a target grayscale image of the surface to be detected of the target furniture, and determining a first number of other pixels in the target grayscale image that have the same grayscale value as the target pixel; determining a first parameter value of the target pixel according to the grayscale values ​​of other pixels in the neighborhood of the target pixel and the first number; determining a gradient feature difference value between the target pixel and other pixels in the neighborhood, and determining a second parameter value of the target pixel according to the first parameter value and the gradient feature difference value; performing saliency detection on the target grayscale image using the second parameter value of the target pixel, obtaining a salient area in the target grayscale image, and using the salient area as the scratch area of ​​the surface to be detected. The method in patent CN118823029B focuses on quality detection of furniture surfaces, and cannot perform rapid detection of furniture quality for rapidly developing and changing furniture product structures. Summary of the invention

[0004] The purpose of this application is to provide a method and device for detecting furniture structure quality based on image recognition, which solves the technical problem of not being able to quickly detect furniture quality for rapidly developing and changing furniture product structures, and achieves the technical effect of quickly detecting furniture quality for rapidly developing and changing furniture product structures.

[0005] An embodiment of the present application provides a furniture structure quality detection method based on image recognition, the method comprising: obtaining multiple sample furniture images of multiple furniture products, and dividing the sample furniture images into multiple sub-sample area images, and dividing the multiple sub-sample area images into multiple sub-sample groups; wherein the sub-sample area images include area images of hole positions, frame corners and frame connections of the furniture products; annotating the sub-sample area images in the multiple sub-sample groups, and training sub-detection systems based on the annotated sub-sample area images in the sub-sample groups; determining multiple sub-detection systems corresponding to the furniture structure according to the furniture structure, and deploying the multiple sub-detection systems corresponding to the furniture structure, wherein the multiple sub-detection systems are used to respectively detect the product quality of the hole positions, frame corners and frame connections of the furniture structure.

[0006] In a possible implementation, a plurality of sample furniture images of a plurality of furniture products are obtained, and the sample furniture images are segmented into a plurality of sub-sample area images, and the plurality of sub-sample area images are divided into a plurality of sub-sample groups, including: obtaining a plurality of sample furniture images of planar frame structures of a plurality of furniture products, and segmenting the sample furniture images into a plurality of sub-sample area images according to the positions of holes, frame corners, and frame connections of the planar frame structures of the furniture products; and dividing the plurality of sub-sample area images into a plurality of sub-sample groups according to the similarities of the hole positions, frame corners, and frame connections.

[0007] In another possible implementation, a plurality of sample furniture images of a plurality of furniture products are obtained, and the sample furniture images are segmented into a plurality of sub-sample area images, and the plurality of sub-sample area images are divided into a plurality of sub-sample groups, further comprising: determining the structural similarity of the planar frame structures of the plurality of furniture products, segmenting the sample furniture images into a plurality of sub-sample area images according to the structural similarity, the plurality of sub-sample area images respectively corresponding to the positions of frame corners and frame connections in the sample furniture images; and dividing the plurality of sub-sample area images into a plurality of sub-sample groups according to the similarities of the frame corners and the frame connections.

[0008] In another possible implementation, multiple sub-detection systems corresponding to the furniture structure are determined according to the furniture structure, including: determining a configuration file corresponding to the furniture structure according to a furniture structure image of the furniture structure through a furniture structure recognition model; and determining multiple sub-detection systems corresponding to the furniture structure according to the configuration file.

[0009] In another possible implementation, the method also includes: displaying a furniture structure image of the furniture structure and multiple sub-detection systems corresponding to the furniture structure, and displaying the correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems, and manually verifying the correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems.

[0010] In another possible implementation, the method also includes: when multiple sub-detection systems respectively detect the product quality of the hole positions, frame corners and frame connections of the furniture structure, outputting the correspondence between the furniture structure image of the furniture structure and the detection results of the multiple sub-detection systems corresponding to the furniture structure; and verifying the correspondence between the furniture structure image of the furniture structure and the detection results of the multiple sub-detection systems corresponding to the furniture structure based on the image area of ​​the manually verified furniture structure image and the correspondence between the multiple sub-detection systems by means of the detection verification unit.

[0011] In another possible implementation, the correspondence between the furniture structure image of the furniture structure and the detection results of the multiple sub-detection systems corresponding to the furniture structure is verified by a detection and verification component according to the correspondence between the image area of ​​the manually verified furniture structure image and the multiple sub-detection systems, including: determining the first area coordinates of the detection result of the first sub-detection system, and determining the second area coordinates of the first sub-detection system of the manually verified furniture structure image, when the distance between the first area coordinates and the second area coordinates is greater than or equal to the preset distance, it is determined that a detection failure has occurred in the first sub-detection system, and a prompt is given to repair the first sub-detection system; when the distance between the first area coordinates and the second area coordinates is less than the preset distance, it is determined that the first sub-detection system is operating normally.

[0012] In another possible implementation, the correspondence between the furniture structure image of the furniture structure and the detection results of the multiple sub-detection systems corresponding to the furniture structure is verified based on the correspondence between the image area of ​​the manually verified furniture structure image and the multiple sub-detection systems, and also includes: obtaining multiple output result area coordinates of the multiple output result areas of the multiple sub-detection systems, and determining the output result distribution characteristics of the multiple output result area coordinates, and obtaining multiple standard result area coordinates of the multiple standard result areas of the manually verified multiple sub-detection systems, and determining the standard result distribution characteristics of the multiple standard result area coordinates; when the similarity between the output result distribution characteristics and the standard result distribution characteristics is greater than or equal to the preset similarity, it is judged that the multiple sub-detection systems are operating normally; when the similarity between the output result distribution characteristics and the standard result distribution characteristics is less than the preset similarity, it is judged that the multiple sub-detection systems are not operating normally, and a prompt is given to maintain the multiple sub-detection systems.

[0013] An embodiment of the present application also provides a furniture structure quality detection device based on image recognition, comprising a unit for executing any of the methods described above.

[0014] An embodiment of the present application also provides a furniture structure quality detection device based on image recognition, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, it implements any of the methods described above.

[0015] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any one of the above items is implemented.

[0016] An embodiment of the present application also provides a computer program product, including a computer program, which implements the steps of any of the methods described above when executed by a processor.

[0017] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0018] The embodiment of the present application provides a furniture structure quality detection method based on image recognition, and the method includes: obtaining multiple sample furniture images of multiple furniture products, and dividing the sample furniture images into multiple sub-sample area images, and dividing the multiple sub-sample area images into multiple sub-sample groups; wherein the sub-sample area images include area images of hole positions, frame corners and frame connections of furniture products; annotating the sub-sample area images in the multiple sub-sample groups, and training sub-detection systems according to the annotated sub-sample area images in the sub-sample groups; determining multiple sub-detection systems corresponding to the furniture structure according to the furniture structure, and deploying the multiple sub-detection systems corresponding to the furniture structure, wherein the multiple sub-detection systems are used to respectively detect the product quality of the hole positions, frame corners and frame connections of the furniture structure. The method in the embodiment of the present application can adapt to the rapid development of small furniture, can quickly configure a device capable of high-quality structural detection for small furniture, improves the quality detection efficiency and configuration efficiency of small furniture, and improves production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0020] Figure 1 A flowchart of a furniture structure quality detection method based on image recognition provided in an embodiment of the present application;

[0021] Figure 2 A schematic diagram of a sub-sample area image obtained by a furniture structure quality detection method based on image recognition in an embodiment of the present application;

[0022] Figure 3 A schematic diagram of the working process of a furniture structure quality inspection method based on image recognition in an embodiment of the present application during quality inspection;

[0023] Figure 4 A flowchart of another furniture structure quality detection method based on image recognition provided in an embodiment of the present application;

[0024] Figure 5 A schematic diagram of the logical structure of a furniture structure quality detection device based on image recognition provided in an embodiment of the present application;

[0025] Figure 6 A schematic diagram of the physical structure of a furniture structure quality detection device based on image recognition provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0027] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0028] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0029] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0030] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0031] Existing image recognition technology cannot adapt to the rapid changes in the structural design of small furniture when conducting furniture quality inspection, and cannot quickly detect the rapidly developing and changing product structure of furniture.

[0032] Based on the above reasons, the embodiment of the present application provides a furniture structure quality detection method based on image recognition, and the method includes: obtaining multiple sample furniture images of multiple furniture products, and dividing the sample furniture images into multiple sub-sample area images, and dividing the multiple sub-sample area images into multiple sub-sample groups; wherein the sub-sample area images include area images of hole positions, frame corners and frame connections of furniture products; annotating the sub-sample area images in the multiple sub-sample groups, and training sub-detection systems according to the sub-sample area images in the annotated sub-sample groups; determining multiple sub-detection systems corresponding to the furniture structure according to the furniture structure, and deploying the multiple sub-detection systems corresponding to the furniture structure, wherein the multiple sub-detection systems are used to respectively detect the product quality of the hole positions, frame corners and frame connections of the furniture structure. The method in the embodiment of the present application can adapt to the rapid development of small furniture, can quickly configure a device capable of high-quality structural detection for small furniture, improves the quality detection efficiency and configuration efficiency of small furniture, and improves production efficiency.

[0033] In some scenarios, a furniture structure quality inspection method based on image recognition of an embodiment of the present application can be applied to the quality inspection of small furniture. It can quickly configure corresponding quality inspection devices in the production of small furniture and can adapt to the rapid iteration and product upgrade of small furniture products.

[0034] The following is a detailed description of a furniture structure quality detection method based on image recognition provided in an embodiment of the present application with reference to specific examples.

[0035] Figure 1 A flowchart of a furniture structure quality detection method based on image recognition is provided in an embodiment of the present application, such as Figure 1 As shown, the method includes S110 to S130, and S110 to S130 are described in detail below.

[0036] S110, obtaining a plurality of sample furniture images of a plurality of furniture products, and dividing the sample furniture images into a plurality of sub-sample area images, and dividing the plurality of sub-sample area images into a plurality of sub-sample groups, wherein the sub-sample area images include area images of hole positions, frame corners, and frame connections of the furniture products.

[0037] In the embodiment of the method, when small furniture products are rapidly iterated, the overall structure of the small furniture products often changes, and local areas of the small furniture products are often relatively similar. In order to improve the adaptability to small furniture products, the small furniture products are split into multiple reusable image detection areas for quality detection, thereby improving the adaptability to the rapid iteration of small furniture products and improving the effect of detecting the structure of small furniture products.

[0038] In an embodiment of the method, a plurality of sample furniture images of a plurality of furniture products can be first obtained, wherein the plurality of furniture products are furniture products from different product series, and the plurality of sample furniture images are furniture images from different furniture product groups. Then, an image detection model for quality inspection of furniture products can be trained using the plurality of sample furniture images, thereby achieving efficient inspection of furniture products.

[0039] After obtaining sample furniture images from different furniture products as a whole, the sample furniture images can be segmented into multiple sub-sample area images, each of which is a local area of ​​the sample furniture image. The sub-sample area images can be used to detect local quality features of the sample furniture image, and then the multiple sub-sample area images can be divided into multiple sub-sample groups. The images of each sub-sample group are images with similar features, and then the quality of different areas of the furniture product can be detected through the sample furniture images of multiple sub-sample groups.

[0040] Figure 2 Schematic diagram of a sub-sample area image obtained by a furniture structure quality detection method based on image recognition in an embodiment of the present application, such as Figure 2 As shown, the multiple sub-sample area images may be sub-sample area image A1 , sub-sample area image A2 , sub-sample area image A3 , and sub-sample area image A4 on the sample furniture image 1 .

[0041] For example, Figure 2 As shown, the sub-sample area images may include area images of hole positions, frame corners and frame connections of furniture products, and product quality inspections may be performed on the hole positions, frame corners and frame connections of furniture products respectively through the sub-sample area images.

[0042] S120: labeling sub-sample region images in a plurality of sub-sample groups, and training a sub-detection system according to the labeled sub-sample region images in the sub-sample groups.

[0043] After obtaining multiple sub-sample groups, the sub-sample area images in the multiple sub-sample groups can be labeled, thereby forming a data set for training an image recognition model. Subsequently, a sub-detection system can be trained based on the sub-sample area images in the labeled sub-sample groups. The sub-detection system is used to perform image recognition on the sub-sample area images of furniture products to detect the structural quality of the furniture products.

[0044] Exemplarily, after labeling the subsample area images in multiple subsample groups, sub-detection systems can be trained separately according to the subsample area images in the multiple subsample groups, and each sub-detection system is used to perform furniture quality detection on the subsample area images in each subsample group.

[0045] S130. Determine multiple sub-detection systems corresponding to the furniture structure according to the furniture structure, and deploy the multiple sub-detection systems corresponding to the furniture structure, wherein the multiple sub-detection systems are used to respectively detect the product quality of the hole positions, frame corners and frame connections of the furniture structure.

[0046] After obtaining multiple sub-detection systems, when it is necessary to perform quality inspection on furniture products with new structural designs, multiple sub-detection systems corresponding to the furniture structure can be determined based on the furniture structure, and then quality inspection of the furniture products can be performed based on the multiple sub-detection systems.

[0047] After determining multiple sub-detection systems, multiple sub-detection systems corresponding to the furniture structure can be deployed. The multiple sub-detection systems are used to respectively detect the product quality of the holes, frame corners and frame connections of the furniture structure, so that the multiple sub-detection systems can be used for quality inspection of different local areas of the furniture products on the production line.

[0048] For example, Figure 3 FIG. 1 is a schematic diagram of a furniture structure quality detection method based on image recognition in an embodiment of the present application during quality detection. Figure 3 As shown, each sub-detection system 2 may include a camera 21 and a computing component 22, wherein the camera 21 is used to capture images of local areas of the furniture products 31 on the conveyor line 32, and the computing component 22 is used to perform quality detection on local areas of the furniture products 31 based on the captured images of the local areas of the furniture products.

[0049] The beneficial effect of the above-mentioned implementation method is that when the rapid iteration of the design of a small furniture product causes changes in the overall structure of the furniture product, and the hole positions, frame corners and frame connections in local areas of the small furniture product are similar, the regional images of the hole positions, frame corners and frame connections of the furniture product can be reused to perform quality inspection on the rapidly iterated furniture products, thereby improving the efficiency of quality inspection of furniture products.

[0050] The beneficial effect of the above-mentioned implementation method is that by repeatedly reusing the area images of the hole positions, frame corners and frame connections of the furniture products to perform quality inspection on rapidly iterating furniture products, the problem of small sample size of furniture product images is solved, making it easier to perform quality inspection on furniture products through image recognition methods.

[0051] The beneficial effect of the above-mentioned implementation method is that, after determining multiple sub-detection systems, it is convenient to deploy multiple sub-detection systems corresponding to the furniture structure. When the product line of furniture products is changed, it is convenient to quickly deploy the quality inspection device of furniture products, thereby improving the quality inspection effect when the home products are changed.

[0052] In some implementations, in the above-mentioned S110, multiple sample furniture images of multiple furniture products are obtained, and the sample furniture images are divided into multiple sub-sample area images, and the multiple sub-sample area images are divided into multiple sub-sample groups, including S111 to S112. S111 to S112 are specifically described below.

[0053] S111, obtaining a plurality of sample furniture images of the plane frame structures of a plurality of furniture products, and dividing the sample furniture images into a plurality of sub-sample area images according to the positions of the holes, frame corners and frame connections of the plane frame structures of the furniture products.

[0054] When determining multiple sub-sample area images, multiple sample furniture images of the plane frame structures of multiple furniture products can be first obtained, and the sample furniture images can be divided into multiple sub-sample area images according to the positions of holes, frame corners and frame connections of the plane frame structures of the furniture products, and then furniture quality inspection can be performed based on the multiple sub-sample area images.

[0055] Exemplarily, in order to ensure the accuracy of the image of furniture products and reduce the image distortion of furniture products, the furniture products can be first disassembled into a planar frame structure, and multiple sample furniture images of the planar frame structures of multiple furniture products can be used as the training data source for the quality inspection model, thereby ensuring the furniture quality inspection effect.

[0056] S112 , dividing the multiple sub-sample area images into multiple sub-sample groups according to the similarities of the hole positions, frame corners, and frame connections.

[0057] After obtaining a plurality of sub-sample area images, the plurality of sub-sample area images may be divided into a plurality of sub-sample groups according to similarities of hole positions, frame corners, and frame connections.

[0058] Exemplarily, by grouping multiple sub-sample area images, the sub-sample area images corresponding to the hole positions of the furniture product can be grouped into a first sub-sample group, the sub-sample area images corresponding to the frame corners of the furniture product can be grouped into a second sub-sample group, and the sub-sample area images corresponding to the frame connections of the furniture product can be grouped into a third sub-sample group.

[0059] Exemplarily, when a plurality of sub-sample region images are divided into a plurality of sub-sample groups, the sub-sample region images may be rotated so that the directions of the sub-sample region images are aligned, which facilitates subsequent training of the image recognition model.

[0060] The beneficial effect of the above-mentioned implementation method is that by performing quality inspection based on multiple sample furniture images of the planar frame structures of multiple furniture products, the accuracy of the images of the furniture products can be ensured and the errors caused by the image distortion of the furniture products can be reduced, thereby ensuring the accuracy of the quality inspection of the furniture products.

[0061] The beneficial effect of the above-mentioned implementation method is that small furniture is often shipped in a disassembled structure when sold online. By first disassembling the furniture products into a flat frame structure and then conducting quality inspection on the furniture, the quality inspection of the furniture can be carried out according to the shipping status, thus ensuring the quality inspection effect of the furniture products.

[0062] In some implementations, in the above S110, multiple sample furniture images of multiple furniture products are obtained, and the sample furniture images are divided into multiple sub-sample area images, and the multiple sub-sample area images are divided into multiple sub-sample groups, and also includes S113 to S114, and S113 to S114 are specifically described below.

[0063] S113, determining the structural similarity of the planar frame structures of the plurality of furniture products, and dividing the sample furniture image into a plurality of sub-sample area images according to the structural similarity, wherein the plurality of sub-sample area images respectively correspond to the positions of the frame corners and the frame connections in the sample furniture image.

[0064] When conducting quality inspections on similar furniture products, such as furniture products from the same series, in order to increase the area of ​​the local region of the inspected furniture products, the area of ​​the local region of the inspected furniture products can be expanded according to the similarity of the furniture products, thereby improving the efficiency of quality inspections on the furniture products.

[0065] When expanding the area of ​​the local region of the detected furniture product, the structural similarity of the planar frame structures of multiple furniture products can be determined, and the sample furniture image can be divided into multiple sub-sample area images according to the structural similarity, each sub-sample area image corresponding to a position with higher structural similarity of multiple furniture products.

[0066] Exemplarily, when the sample furniture image is segmented into a plurality of sub-sample region images according to structural similarity, the sample furniture image corresponding to a position with a structural similarity of 100% (ie, the structure is completely the same) may be segmented into sub-sample region images.

[0067] Exemplarily, when the sample furniture image is segmented into a plurality of sub-sample region images according to structural similarity, the sample furniture image corresponding to positions having the same structural shape but different sizes may be segmented into sub-sample region images.

[0068] Exemplarily, a plurality of sub-sample area images may respectively correspond to the positions of frame corners and frame connections in the sample furniture image, and then quality inspection of the frame corners and frame connections of the furniture product may be performed based on the plurality of sub-sample area images.

[0069] S114 , dividing the multiple sub-sample area images into multiple sub-sample groups according to the similarities of the frame corners and the frame connections.

[0070] After obtaining multiple subsample area images, the multiple subsample area images can be divided into multiple subsample groups according to the similarities of frame corners and frame connections, and then the quality of furniture products can be inspected by using the subsample groups including the multiple subsample area images.

[0071] Exemplarily, when respectively determining the similarities of the frame corners and the frame connections, the similarities of the frame corners may be determined according to the size and shape of the frame corners, and the similarities of the frame connections may be determined according to the size and shape of the frame connections.

[0072] The beneficial effect of the above-mentioned implementation method is that the structural similarity of the planar frame structures of multiple furniture products is determined, and the sample furniture image is divided into multiple sub-sample area images according to the structural similarity, which can expand the area of ​​the region for quality inspection of furniture products and improve the efficiency of quality inspection of furniture products.

[0073] The beneficial effect of the above-mentioned implementation method is that, according to the similarity of the frame corners and the frame connections, multiple sub-sample area images are divided into multiple sub-sample groups, which can ensure the training effect of the image detection model and ensure the effect of quality inspection of furniture products by sub-sample groups of multiple sub-sample area images.

[0074] In some implementations, in the above S130, multiple sub-detection systems corresponding to the furniture structure are determined according to the furniture structure, including S131 to S132, and S131 to S132 are specifically described below.

[0075] S131. Determine a configuration file corresponding to the furniture structure according to the furniture structure image of the furniture structure by using a furniture structure recognition model.

[0076] In an embodiment of the present application, when it is necessary to perform quality inspection on furniture products with new structures, in order to quickly determine the sub-inspection system for quality inspection of the furniture products, a furniture structure recognition model can be used to determine the configuration file corresponding to the furniture structure based on the furniture structure image of the furniture structure. The configuration file corresponding to the furniture structure is used to determine multiple sub-inspection systems corresponding to the furniture structure.

[0077] Exemplarily, the furniture structure recognition model can be obtained by training the furniture structure image and a plurality of sub-detection systems corresponding to the furniture structure image.

[0078] Exemplarily, the configuration file may be a text file that records information of multiple sub-detection systems.

[0079] S132. Determine multiple sub-detection systems corresponding to the furniture structure according to the configuration file.

[0080] After obtaining the configuration file, multiple sub-detection systems corresponding to the furniture structure can be determined according to the configuration file, and then the quality of the furniture structure of the new structure can be detected according to the multiple sub-detection systems.

[0081] The beneficial effect of the above implementation method is that by determining multiple sub-detection systems corresponding to the furniture structure through the furniture structure recognition model, efficient quality detection of furniture products with new structures can be achieved.

[0082] The beneficial effect of the above implementation is that the information of multiple sub-detection systems is recorded in the configuration file, thereby improving the processing efficiency of the information of multiple sub-detection systems.

[0083] In some implementations, the above method also includes: displaying a furniture structure image of the furniture structure and multiple sub-detection systems corresponding to the furniture structure, and displaying the correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems, and manually verifying the correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems.

[0084] After determining the multiple sub-detection systems corresponding to the furniture structure image, the furniture structure image of the furniture structure and the multiple sub-detection systems corresponding to the furniture structure can be displayed, so as to realize real-time display of the furniture structure image and the multiple sub-detection systems, thereby facilitating real-time display of the quality inspection results of the furniture products and facilitating manual verification of the inspection results of the multiple sub-detection systems.

[0085] After displaying the furniture structure image of the furniture structure and the multiple sub-detection systems corresponding to the furniture structure, the correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems can be displayed. The correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems is the position correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems, which makes it easy to manually verify the correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems to ensure the accuracy of the detection areas of the multiple sub-detection systems.

[0086] Exemplarily, when displaying the correspondence between the image area of ​​the furniture structure image and multiple sub-detection systems, the image area of ​​the furniture structure image and the multiple sub-detection systems can be connected by lines to intuitively display the correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems.

[0087] The beneficial effect of the above implementation method is that by intuitively displaying the correspondence between the image area of ​​the furniture structure image and multiple sub-detection systems, it is convenient to manually verify the position correspondence between the image area detections of the furniture structure image by multiple sub-detection systems.

[0088] The beneficial effect of the above-mentioned implementation method is that, by intuitively displaying the correspondence between the image area of ​​the furniture structure image and multiple sub-detection systems through lines, it is convenient to arrange multiple subsystems according to the image area of ​​the furniture structure image, thereby improving the efficiency of arranging the quality inspection system for furniture products.

[0089] Figure 4 A flowchart of another furniture structure quality detection method based on image recognition provided in an embodiment of the present application is shown in FIG. Figure 4 As shown, the above method further includes S210 to S220, and S210 to S220 are described in detail below.

[0090] S210, when multiple sub-detection systems respectively detect the product quality of the hole positions, frame corners and frame connections of the furniture structure, output the correspondence between the furniture structure image of the furniture structure and the detection results of the multiple sub-detection systems corresponding to the furniture structure.

[0091] When conducting quality inspection on furniture products, multiple sub-inspection systems can be used to respectively inspect the product quality of the holes, frame corners and frame connections of the furniture structure, and the inspection results of the multiple sub-inspection systems can be obtained. The inspection results of the multiple sub-inspection systems can be directly displayed on the furniture structure image, that is, the correspondence between the furniture structure image of the furniture structure and the inspection results of the multiple sub-inspection systems corresponding to the furniture structure can be directly output, thereby facilitating the verification of the inspection results of the multiple sub-inspection systems.

[0092] Exemplarily, the detection result of the sub-detection system can be directly displayed on the furniture structure image by marking the area on the furniture structure image through a label.

[0093] S220, verifying the correspondence between the furniture structure image of the furniture structure and the detection results of the multiple sub-detection systems corresponding to the furniture structure based on the correspondence between the image area of ​​the manually verified furniture structure image and the multiple sub-detection systems through the detection and verification unit.

[0094] When verifying the detection results of multiple sub-detection systems, the detection verification unit can verify the correspondence between the furniture structure image of the furniture structure and the detection results of multiple sub-detection systems corresponding to the furniture structure based on the correspondence between the image area of ​​the manually verified furniture structure image and the multiple sub-detection systems, so as to achieve accurate verification of the detection results of multiple sub-detection systems.

[0095] During operation, the correspondence between the image area of ​​the manually verified furniture structure image and the multiple sub-detection systems is the accurate position of the detection results that the multiple sub-detection systems should output. Then, the correspondence between the furniture structure image of the furniture structure and the detection results of the multiple sub-detection systems corresponding to the furniture structure can be verified through the accurate position of the detection results that should be output, so as to avoid the deviation of the detection positions of the multiple sub-detection systems and ensure the normal working state of the sub-detection systems.

[0096] The beneficial effect of the above implementation method is that by verifying the actual detection results of multiple sub-detection systems through the accurate positions of the detection results to be output, the detection positions of multiple sub-detection systems can be avoided from being offset, and the normal working state of the sub-detection systems can be ensured.

[0097] In some implementations, in the above S210, the corresponding relationship between the furniture structure image of the furniture structure and the detection results of the multiple sub-detection systems corresponding to the furniture structure is verified by the detection and verification component according to the corresponding relationship between the image area of ​​the manually verified furniture structure image and the multiple sub-detection systems, including: determining the first area coordinates of the detection result of the first sub-detection system, and determining the second area coordinates of the first sub-detection system of the manually verified furniture structure image, when the distance between the first area coordinates and the second area coordinates is greater than or equal to the preset distance, it is determined that the first sub-detection system has a detection failure, and a prompt is given to maintain the first sub-detection system. When the distance between the first area coordinates and the second area coordinates is less than the preset distance, it is determined that the first sub-detection system is operating normally.

[0098] When the actual detection results of multiple sub-detection systems are verified by the accurate position of the detection results to be output, the first area coordinates of the detection results of the first sub-detection system can be determined, and the second area coordinates of the first sub-detection system of the manually verified furniture structure image can be determined, and then the actual detection results of multiple sub-detection systems can be verified according to the accurate position of the detection results to be output by the first area coordinates and the second area coordinates.

[0099] When comparing the actual detection results of multiple sub-detection systems with the accurate positions of the detection results that should be output, when the distance between the first area coordinates and the second area coordinates is greater than or equal to the preset distance, it means that a large deviation has occurred in the actual output result position of the first sub-detection system, and thus it can be determined that a detection failure has occurred in the first sub-detection system, prompting maintenance of the first sub-detection system and adjustment of the first sub-detection system.

[0100] Exemplarily, when the first sub-detection system is maintained, the detection position of the camera of the first sub-detection system may be adjusted.

[0101] When comparing the actual detection results of multiple sub-detection systems with the accurate positions of the detection results to be output, when the distance between the first area coordinates and the second area coordinates is less than the preset distance, it means that there is no large deviation in the actual output result position of the first sub-detection system, and thus it can be determined that the first sub-detection system is operating normally.

[0102] The beneficial effect of the above-mentioned implementation method is that by determining the first area coordinates of the detection results of the first sub-detection system and determining the second area coordinates of the first sub-detection system of the manually verified furniture structure image, it is possible to accurately verify the position of the actual detection results of the first sub-detection system, thereby improving the efficiency of monitoring the working status of multiple sub-detection systems.

[0103] In some implementations, in the above S210, the correspondence between the furniture structure image of the furniture structure and the detection results of multiple sub-detection systems corresponding to the furniture structure is verified based on the correspondence between the image area of ​​the manually verified furniture structure image and the multiple sub-detection systems, and also includes S211 to S212. S211 to S212 are specifically explained below.

[0104] S211. Obtain multiple output result area coordinates of multiple output result areas of multiple sub-detection systems, and determine the output result distribution characteristics of the multiple output result area coordinates, and obtain multiple standard result area coordinates of multiple standard result areas of multiple sub-detection systems that are manually verified, and determine the standard result distribution characteristics of the multiple standard result area coordinates.

[0105] When verifying the positions of actual output results of multiple sub-detection systems, multiple output result area coordinates of multiple output result areas of the multiple sub-detection systems can be obtained, and output result distribution characteristics of the multiple output result area coordinates can be determined, and the output result distribution characteristics represent the position distribution characteristics of the actual output results of the multiple sub-detection systems.

[0106] When verifying the positions of actual output results of multiple sub-detection systems, it is also possible to obtain multiple standard result area coordinates of multiple standard result areas of multiple sub-detection systems that are manually verified, and determine the standard result distribution characteristics of the multiple standard result area coordinates. The standard result distribution characteristics characterize the position distribution characteristics of the target output results of the multiple sub-detection systems.

[0107] S212: When the similarity between the output result distribution feature and the standard result distribution feature is greater than or equal to the preset similarity, it is determined that the multiple sub-detection systems are operating normally. When the similarity between the output result distribution feature and the standard result distribution feature is less than the preset similarity, it is determined that the multiple sub-detection systems are not operating normally, and a prompt is given to perform maintenance on the multiple sub-detection systems.

[0108] When verifying the positions of actual output results of multiple sub-detection systems, when the similarity between the output result distribution characteristics and the standard result distribution characteristics is greater than or equal to the preset similarity, the multiple sub-detection systems are judged to be operating normally and there is no need to adjust the multiple sub-detection systems.

[0109] When verifying the positions of actual output results of multiple sub-detection systems, when the similarity between the output result distribution characteristics and the standard result distribution characteristics is less than the preset similarity, it is determined that the multiple sub-detection systems are not operating normally, and a prompt is given to perform maintenance on the multiple sub-detection systems. The multiple sub-detection systems need to be adjusted to ensure the stability of the working status of the multiple sub-detection systems.

[0110] Exemplarily, when verifying the positions of actual output results of multiple sub-detection systems, the output result distribution characteristics and the standard result distribution characteristics can be respectively the distance matrices between the multiple detection results, and then the similarity of the distance matrices corresponding to the output result distribution characteristics and the standard result distribution characteristics can be determined as the similarity between the output result distribution characteristics and the standard result distribution characteristics.

[0111] The beneficial effect of the above implementation is that the positions of the output results of multiple sub-detection systems are verified by the similarity between the output result distribution characteristics and the standard result distribution characteristics, thereby improving the accuracy of verifying the working status of multiple sub-detection systems.

[0112] An embodiment of the present application also provides a furniture structure quality detection device based on image recognition, comprising a unit for executing any of the methods described above.

[0113] Figure 5 A schematic diagram of the logical structure of a furniture structure quality detection device based on image recognition provided in one embodiment of the present application is shown in FIG. Figure 5 As shown, the device 4 of this embodiment includes a processing unit 41, a storage unit 42 and a transceiver unit 43. The processing unit 41 is used to process data, the storage unit 42 is used to store data, and the transceiver unit 43 is used to send and receive data. The processing unit 41, the storage unit 42 and the transceiver unit 43 cooperate with each other to implement the above method. The beneficial effects of the embodiment of the present application have been described in the above method and will not be repeated here.

[0114] An embodiment of the present application also provides a furniture structure quality detection device based on image recognition, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, it implements any of the methods described above.

[0115] Figure 6 A schematic diagram of the physical structure of a furniture structure quality detection device based on image recognition provided in one embodiment of the present application, such as Figure 6 As shown, the device 5 of this embodiment includes: at least one processor 50 ( Figure 6 Only one processor 50 is shown in the figure), a memory 51, and a computer program 52 stored in the memory 51 and executable on the at least one processor 50. When the processor 50 executes the computer program 52, the steps in any of the above-mentioned method embodiments are implemented. The beneficial effects of the embodiments of the present application have been described in the above-mentioned methods and will not be repeated here.

[0116] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0117] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0118] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0119] An embodiment of the present application provides a computer program product. When the computer program product runs on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0120] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0121] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0122] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0123] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0124] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0125] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A furniture structure quality detection method based on image recognition, characterized in that: The method comprises: Acquire multiple sample furniture images of multiple furniture products, and divide the sample furniture images into multiple sub-sample area images, and divide the multiple sub-sample area images into multiple sub-sample groups; wherein the sub-sample area images include area images of hole positions, frame corners, and frame connections of the furniture products; Annotating sub-sample area images in a plurality of sub-sample groups, and training a sub-detection system based on the annotated sub-sample area images in the sub-sample groups; According to the furniture structure, multiple sub-detection systems corresponding to the furniture structure are determined and deployed. The multiple sub-detection systems are used to respectively detect the product quality of the hole positions, frame corners and frame connections of the furniture structure.

2. The method according to claim 1, characterized in that Acquiring a plurality of sample furniture images of a plurality of furniture products, and segmenting the sample furniture images into a plurality of sub-sample area images, and dividing the plurality of sub-sample area images into a plurality of sub-sample groups, including: Acquire multiple sample furniture images of planar frame structures of multiple furniture products, and segment the sample furniture images into multiple sub-sample area images according to the positions of holes, frame corners, and frame connections of the planar frame structures of the furniture products; According to the similarity of hole positions, frame corners and frame connections, multiple sub-sample area images are divided into multiple sub-sample groups.

3. The method according to claim 2, characterized in that Acquiring a plurality of sample furniture images of a plurality of furniture products, dividing the sample furniture images into a plurality of sub-sample area images, and dividing the plurality of sub-sample area images into a plurality of sub-sample groups, further comprising: Determine the structural similarity of the plane frame structures of multiple furniture products, and divide the sample furniture image into multiple sub-sample area images according to the structural similarity, wherein the multiple sub-sample area images respectively correspond to the positions of the frame corners and the frame connections in the sample furniture image; According to the similarities of frame corners and frame connections, multiple sub-sample area images are divided into multiple sub-sample groups.

4. The method according to claim 3, characterized in that According to the furniture structure, multiple sub-detection systems corresponding to the furniture structure are determined, including: Determine, by means of a furniture structure recognition model, a configuration file corresponding to the furniture structure according to the furniture structure image of the furniture structure; Multiple sub-detection systems corresponding to the furniture structure are determined according to the configuration file.

5. The method according to claim 4, characterized in that The method further comprises: Display the furniture structure image of the furniture structure and multiple sub-detection systems corresponding to the furniture structure, and display the correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems, and manually verify the correspondence between the image area of ​​the furniture structure image and the multiple sub-detection systems.

6. The method according to claim 5, characterized in that The method further comprises: When multiple sub-detection systems detect the product quality of the hole positions, frame corners and frame connections of the furniture structure respectively, the corresponding relationship between the furniture structure image and the detection results of the multiple sub-detection systems corresponding to the furniture structure is output; The detection and verification unit verifies the correspondence between the furniture structure image of the furniture structure and the detection results of the multiple sub-detection systems corresponding to the furniture structure based on the correspondence between the image area of ​​the manually verified furniture structure image and the multiple sub-detection systems.

7. A furniture structure quality detection device based on image recognition, characterized in that: Comprising means for performing the method according to any one of claims 1 to 6.

8. A furniture structure quality detection device based on image recognition, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images